ETHIndia 2022 - AMA with Vitalik Buterin

All right. Welcome everybody. Who is excited? There we go. I'm Kartik.

With me I have Juan Benet from Protocol Labs and we are going to be your moderators for the next hour. We're going to talk about a handful of things. We sourced a lot of questions from from the audience here. We have a lot of tweets that we're going to have him justify and overall want to just learn about what he's thinking about and ask some of the questions that you submitted. So, please join me in welcoming Vitalik Buterin from the Ethereum Foundation.

Woo! All right, we're going to do a quick mic test Vitalik. How's it going? I I think I think I'm good. I don't know.

Are you see Are you hearing me well? Well, we can hear you. We just can't understand what you said. Try one more time. Okay, hold on.

Um 1 2 3 4 There we go. We have now successfully calibrated. Well, thank you for joining us. There's a a few thousand people in front of you. We'll kind of switch between the audience here and well, we're going to make the best use of the next 45 minutes and I'm going to start with a few questions that we got.

A lot of them are definitions that people are kind of excited to understand a bit deeper about. So, my first kind of question is a couple of months ago in your talk on Eth Mexico, you talked about how the next step is to reduce the transaction costs and fees to by 100x and a lot of that is just improvements to scalability and how we kind of add ways to compress and and kind of roll up information, but the question that we kind of got from that was could you tell us what gas is as a concept and why is there a gas limit and why is there a fee market, and overall, how does that change over time as we go to super low transaction costs? Um, sure. Uh, so, because it in a blockchain, uh, every full node in the network needs to process, uh, every transaction that it gets included, there has to be some limit to the number of, uh, transactions that get included into the chain, right? Because, uh, if everyone was able to include transactions without limits, then people would, like, run, basically, their web servers on top of it.

Like, I would put an encrypted backup of all my, uh, music and my pictures onto it. And there's just like all kinds of extremely low-value uses that if they were actually allowed on chain, they would add up to like such a huge amount of space that it there would possibly only be like four or five computers in the world that would actually be willing to run a node, and the whole chain would just be super centralized and not very secure, right? So, to prevent that kind of situation and, uh, to uh, strike a balance between, um, making it reasonably viable for people to write to the chain and, uh, making it reasonably viable for people to read the chain, there's this notion of a, uh, transaction fee market, right? Where the two sides of the coin of that are transaction fees and a, uh, block size limit, right? And they're related because, you know, the um, supply and demand, right?

So, so, the higher the uh, block size limit, which is the, uh, parameter that's actually set in the protocol, the lower the fees end up being. And then, if you uh, shrink the limit, then at least in the short term, the fees become higher because, uh, there's just more people clamoring for and competing for a much more limited amount of space, right? And so, every transaction has to specify a fee. It is, uh, Now, originally, it's it was a fairly simple mechanism. The transaction specify a fee, and that's just like an amount of ETH and if it gets included in the chain then it pays that amount.

If it doesn't get included then it does not pay that amount, right? So, gas is a mechanism that is used to measure the amount of resources that is being consumed to buy a transaction, right? So, not all transactions have the same size and you don't want to treat big transactions the same as small transactions because well, big transactions actually have a much higher load on the network that they're imposing and a network that can handle you know, 1,400 small transactions in a block. That's like actually roughly the maximum amount of transactions that can get into a block right now. It can't necessarily handle you know, 1,400 big monster transactions that are trying to you know, prove some complicated ZK-Snark or create a new contract and get it included on chain or like stake 100 kilobytes of data on chain or something like that, right?

And so, gas is an attempt to measure the resources that are being consumed by a transaction. So, there For example, the base cost of a transaction is 21,000 gas and an opcode, so one computational step in the Ethereum Virtual Machine is uh somewhere between two to 10 gas for most operations. So, things like just pushing and popping numbers onto the stack very simple stuff that's stable for gas but then some module or math operation that's a little bit more involved goes up to 10 gas and then there are a few things that take even more gas, right? Um bytes in a transaction. So, every byte in a transaction in the transactions data is 16 gas if it's not a zero and four gas if it is a zero.

Um Editing the storage, so editing the blockchain state is it depends on circumstance but like roughly speaking, it's about 5,000 gas if you're just changing a the value of a storage slot, it's about 2,100 gas if you're just reading, and it's a 20,000 gas if you're filling a new storage slot, right? So, there is like a lot of these numbers in the protocol, but they're all intended to measure like basically how much of a load is your transaction imposing on the Ethereum network, and then the fee that you have to pay is proportional to that, right? So, the limit in the in the chain is is the gas limit. So, there's a limit of 30 million gas that can be spent in a single block, right? So, that's where the 1,400 number comes from.

30 million divided by the 21,000 base cost, right? It's going to be a one about a 1,428. Um but uh And so, that's like, you know, basically the basically the fee market, right? Now, recently there has been this change made called EIP-1559, which kind of changes the mechanics a bit, right? Where there's this concept of a base fee and a priority fee, and that's just I mean, not important to talk about that in too much detail yet, but basically that's just a change that tries to make it more the the fee market more efficient.

So, before one of the flaws that of these system was like, you would send a transaction, you would choose a fee, and then when you choose a fee, you would make this complicated balance between setting a fee higher to make sure that you get in really quickly versus setting a fee lower to save money, and but potentially have to just like wait an extra five or five minutes. But like, sometimes it's 1 minute, sometimes it's 5 minutes, sometimes it's an hour, and it was just very unpredictable. And EIP-1559 basically kind of flipped it around and said, instead of the user setting a fee, the protocol sets the fee, and then the user just has to pay the fee, and they have to set the maximum that they're willing to pay, and they have to set like a little bit extra to as an incentive for the transaction to get included in a block. But now, you know, as a result of that, like basically the mechanism is like you set roughly, you know, you set the max that you're willing to pay, your transaction gets included, you you pay whatever the fee is uh that's uh determined by the protocol, and that goes up and down over um over time, but, you know, at the end of the day it's basically just a mechanism that makes sure that uh you know, these very low value use cases that uh cannot uh like really are not worth it to get included in the chain, right? So, like if, for example, I tried to send millions of uh transactions to upload an encrypted uh copy of my uh music or my photo sets to the uh blockchain, then that would like cost me incredibly huge amounts of money, and it would not be worth it for me, but if I just want to send transactions to do normal stuff, and it's like, you know, a couple hundred bytes to verify a couple signatures, then I'll be able to do that, right?

So, that's uh what the transaction fee system does, and gas is just like the unit of uh computation. That was uh incredibly comprehensive and informative. This is This is awesome. Um I'll I This already kind of transitions into like another question that came in, uh and this was from uh one of the tweets that we saw from an AMA you did a while ago. Um could you give us an update on uh what's happening with state rent, and uh just well, what that problem is, and what the potential solution is, and where is that solution in the implementation?

Mhm. Okay, yes, a very important question. Uh so, there's a few different re- kinds of uh resource one chain can consume, right? So, we talked about computation. If your transaction uh can consumes computational resources, then once that transaction gets included in a block, and every node that has to download and verify that block has to run that computation, and that's a one-time cost for that node, right?

There's data cost, which are the same, except instead of a computational cost, it's data bandwidth. And then there are storage costs. There's two kinds of storage costs. There's the cost of reading and and editing storage, right? So, I mentioned 21,000 2100 gas to read a storage slot, and 5,000 gas to edit a storage slot.

And these are really just a different kind of computation, right? There's just like a one-time thing where instead of like your computer having to do math, your computer has to read uh to or write from the disk, right? Now, there is another kind of storage cost, which is the cost of filling new storage slots. So, creating a new account that did not exist before, filling a new storage slot that before was zero cuz every storage slot is zero by default and now becomes non-zero. Um at creating a new contract that has a big piece of code, right?

So, anything that increases the total size of stuff that goes into what's called the Ethereum state, and that nodes currently have to store forever, right? And the problem with the state is that it is not a one-time cost. It's a permanent cost, right? And uh it's uh permanently increases the amount of work that you know what's up to do. Uh so, right now, I think the state is like somewhere between like 40 and 150 GB, depending on what parts you count, how hard you try to compress it, whether or not you include the Merkel tree.

Um but that's a number that's increasing over time. And that's a cost that we're stuck with, right? So, like even if everyone stops using Ethereum tomorrow, any new node joining the network would still have to download those and many gigabytes pretty much until the end of time, right? And so, this is kind of unfair, right? Because it basically means that you can pay once in order to force the Ethereum network to suffer these costs forever.

So, there's two approaches uh to try to deal with this problem. One of those approaches is to just say, "Well, okay, we're just going to set the gas cost for filling new storage really high, right?" And that can be done, and to some extent, we kind of do it already, right? So, like as I mentioned, there's a cost of 5,000 to edit a slot and 20,000 to create a new slot, but at the same time, there's like a lot of economic problems that appear if you lean on that too hard, right? So, like one of those problems is that, well, there's some applications that really need storage for only a few months and there's some applications that need storage for 30 years and you want to charge more for the second than the first.

Um another challenge is, well, gas prices go up and down a lot, right? Like sometimes the gas price is five gwei. Um a gwei is a millionth of an ETH and uh when we say five gwei or so a billionth of an ETH. And when we say five gwei, that means five billionths of an ETH uh per gas, right? So, like for a simple transaction, that would be five times 21,000 gwei, which is 105,000 gwei, which is uh 0.

000105 ETH. So, it's uh around like 14 cents uh to pay for a transaction for a basic transaction, right? Um so, that would Sometimes when So, when the gas price is five gwei, right? You know, you can multiply that by 20,000. So, you're filling a new storage slot, that's uh 0.

0001 ETH and it will take you about uh 13 cents to fill a storage slot, but then when, let's say uh if instead of uh filling the storage on a Sunday, you want to fill the storage on a Wednesday, well, maybe the gas price will go go up to like 50 gwei and then instead of paying 13 slots, you have to fill it uh pay a dollar 30 to fill a storage slot, right? And this is uh this does not make much economic sense because like really the difference in terms of the cost you're imposing on the network between filling the storage slot on a Wednesday and filling the storage slot on a Sunday is very tiny, right? The difference is just four days and then otherwise the many, many years after that it's basically the same, right? So, it makes no sense that you have to pay 10 times more to fill it on a Wednesday than to fill it on a Sunday, Right? So, that's another problem um and then there is a third problem which is that like it's philosophically ugly because it still means means that storage keeps on increasing forever and it would would be really nice if we could have a balance, right?

That says, well, the maximum storage there is a maximum to the storage that a client a client needs to have and that's going to be proportional to the gas limit. So, there's all these reasons why just charging higher doesn't really work well. And so, the big category of alternative solutions that we've thought about for a long time is well, what if we take away this idea that you can fill a storage slot once and have that be part of the state that everyone needs to have that's accessible forever, right? So, instead, if you fill a storage slot, let's say by default you get it for 6 months and then if you want to have it longer, well, you'll have to pay for it again and again. Right?

So, this is very economically clean and it solves a lot of the problems. If you want us to fix this issue of like, well, make why should it be 10 times more expensive on a Wednesday than on a Sunday than, you know, you can even have a separate a kind of different gas counter that has a separate fee adjustment mechanism that doesn't have a gas limit that kind of accounts for that and so, the price for that could be much more stable than the price for like just computation and bandwidth. So, that's like a very economically nice, but it has this really important problem which is that well, basically it's a UX nightmare, right? Like if you have ERC-20 tokens, does that mean that your ERC-20 tokens disappear every 6 months unless you actively maintain them? What about storage slots or applications that do not correspond to a particular user, right?

So, like let's say if we take an ERC-20 as an example. In an ERC-20, the storage slot that corresponds to your balance doesn't go into your account. It goes into the contract that is the ERC-20 token. And so, the ERC-20 token then needs to have this extra complicated logic to kind of pass on the cost of maintaining that storage slot to the user, right? Or in another one, like what if you someone sends you 100 coins and then you wait for 6 months and then let's say the contract tries to include some mechanism to let you recover those coins, like you prove that you had them at some point in history and then you recover them.

Well, what if before that someone else sent you 50 coins? And now the 100 and the 50 have to be added together, right? So, there's just all of these really complicated developer experience consequences. And about like 1 and 1/2 years ago, I made this proposal. It's the state expiry EIP.

Let's see if I can find it. It's uh Um if you go to nodes.ethereum.org/at the bottom and state_expiry_EIP, I'll just like send that link over there. I think it's like I get I have this concrete proposal for how to do state expiry.

And it's kind of complicated. It's like you can see some of the some of the diagrams down near the bottom. It's you know basically, instead of having one state, you have multiple states and you create a huge tree every year and there's a mechanism for kind of keeping your state updated. There's a mechanism where well, if your state becomes old, then it can't be accessed by the EVM directly, but if you provide a proof that that state existed, then you can revive it. So, there's like a mechanism that's less bad than anything else I've seen anyone come up with, but like you know, it's still kind of annoying, right?

And so, the question is well, are we going to be forced to do that or are we just going to kind of accept the cost of a permanent ever growing state and not deal with that? Now, there I Here we do need to get a bit into the proposer builder separation discussion, right? Because a PBS actually massively decreases the costs of just not caring about the problem and letting the state grow to a terabyte, right? So, basically what happened with PBS is that there's this uh idea of separating the role of creating a block from the role of proposing. Uh so, a proposing but proposing sort of I mean is like being a validator, right?

So, if you sent 32 ETH into the deposit contract, you can be a validator and roughly once every 2 months you get assigned the ability to create a block. And when you create a block, you choose the contents of the block, right? What we've discovered over the last couple of years is this concept of MEV, right? And roughly what this means is that there is this extra revenue that you can get on top of collecting transaction priority fees um that where you collect that revenue like you can only collect that revenue if you are the creator of that block and so you get the ability to choose what are the first transactions that go into the block, right? So, one simple example of this is uh just crypto market arbitrage, right?

So, the uh price of ETH is uh 1272 right now, but let's say uh in the next 12 seconds, some event minor event happens. Could be random market movements, could be um you know, some random news about regulation, could be some centralized exchange going bankrupt, could be lots of things. And let's say on Binance over the next 12 seconds, the uh ETH price drops from 1272 to 1262. Now, on the blockchain, there's still a bunch of decentralized exchanges, other platforms whose last memory is that the Ethereum price is 1272, right? So, if you take Uniswap, for example, well, what could you do then?

What you could do is you could sell some ETH on Uniswap in the first transaction of the next block, and you could keep selling until the price drops from 1272 to 1262, right? And then you would collect some some revenue from that, and then at the same time, you would buy the ETH on Binance at 1262, right? So you're basically arbitrating the two, and on average you're making that difference of $5 of ETH as profit, right? So that is revenue that you can get by being the first person to create or or the person to choose the first transaction in a block. And that's that kind of revenue and continuing to earn that kind of revenue, especially as the ecosystem continues to evolve and there's more exchanges, there's more DeFi, there's more fancy stuff happening on chain, that's like that requires sophisticated algorithms, right?

And that's not accessible to everyone, that involves proprietary algorithms, that requires having a lot of skill. And so the concern was that, well, you know, are there going to be some stakers that are much better at this than other stakers, and are those stakers going to create pools that everyone joins that just end up dominating, right? And we have numbers and the revenue that MEV or stakers that use amazing MEV algorithms get right now is from producing blocks is about double the revenue that regular stakers get, right? So that's not including attestations or any other benefit, that's just like just taking into account proposals. But uh the like this creates centralization pressure, right?

And what we wanted to do to fix that centralization pressure is basically create these open markets where individual stakers can say, well, I'm going to outsource the right to create a block to someone else, and it's going to be an open market, so people who have these sophisticated algorithms can go and create the block, and I'm just going to accept whoever offers me the highest bid, right? And then I, as an individual staker, would be able to get the benefit of earning I think it's like on average about 0.16 ETH instead of 0.08 ETH per block. Um but and then but I personally would not have to like incur the burden of being having all of this specialized software, specialized hardware, specialized equipments to try to like try really hard to collect this extra revenue, right?

So, that So, one of the interesting side effects of that mechanism is that if all I'm doing is accepting bids, then I as a staker eventually am not going to actually have to even store the Ethereum state on my node, right? So, this requires another upgrade which is called stateless clients. Um stateless clients is basically I guess this upgrade that where nodes can verify Ethereum blocks without having the state themselves. Um and the way that this works is basically that whoever creates the block um has the that block it contains its transactions, it contains a the Merkel the root hash of the new state. And and then it also has a list of of keys and values that basically say, "This block touched the code of account number 1 2 3 4 5 and here is the code.

This block touched storage slot number 6 7 8 of account number 9 1 2 and here is the and here it was the previous value. This block block touched the balance of account number 3 4 5 6 and its balance before this was 5,000." So, it just declares what all the states are that it that it read and wrote to and it also includes a proof that those states are actually correct, right? And if you want to learn how this works you can there is a block block post that I have on Merkel trees. So, you know, go to I think it's alex.

ca and just control up for Merkel that's like v e r k l e. That's I'll just link that to the article too but that's how it works, right? But the the property that does is that it basically means that um you know, if you, uh, as a person running a node, I would not have to have any of the state myself, and other people can go have the state uh other the people who create the blocks, or the builders, would be the ones that have to have the state, and they're already more specialized actors, so they're okay having many terabytes of data, and I, as a user as a, uh, build proposer, it can just be very lightweight, even if the state is extremely big, and so that just reduces the cost of the state being big, right? So, to circle back, the debate is basically, you know, given this kind of P proposer builder separation reality, do we even need to do state expiry, or can we just kind of like let the state get big, and accept the fact that there is a smaller number of specialized nodes that are dealing with it? And that's like an ongoing discussion, and state expiry could be something that can be added after 5 or 10 or 20 years if we need to, so it's, uh, especially given PBS, I think it's, uh, I think it's generally agreed by core developers to be a low priority thing, right?

But that's like roughly what the debate is. Okay, fine, Kartik, I see you yelled at me, so I'll try to be faster at the other questions, but Uh, the goal, um, no, that was that was awesome and super comprehensive and super useful. I think a a lot of people learned a lot through that. Um, we have a bunch of questions, so maybe for, uh, the upcoming ones, uh, let's aim for like two two to three minutes, uh, for so that we can get through many, but, uh, that was awesome, uh, and I think a lot of people We'll share the links on the live stream, uh, the the two links that, uh, we just got from him on the referencing the articles, and, uh, we'll we'll get that out there. Yep.

So, um, there's some questions around like public goods funding. I think a lot of people have gotten super interested in the ability to use crypto and web3 to fund public goods. Um, uh, but it might be useful to for those of us that are new to web3 to kind of talk about, uh, a little bit of a high level, why what are public goods? Why are they important? What are some of the mechanisms that you think are super good for funding public goods in web3?

What are some of the achievements that you're, uh, excited about that you think, you you it's great that we've done X, Y, and Z, or something. And maybe one of the experiments that you think are in the horizon that maybe people should think about trying and and who are they might be excited about. A lot of people are hacking now and will be hacking in future hackathons over the next few months. And so, if you kind of see some really good ideas now, they they might get built over time. Yeah, so there is a lot of different ways to define public goods.

And like we can get into the technicalities technicalities of like excludability and rivalrousness and all of those things. But I think the definition that matters the most is like things that are valuable to lots of people, but where there is no good way for the creator to capture that value without you know, completely ruining the the value of the good, right? So, one simple example relevant to the crypto space is uh the design of new algorithms, right? So, you know, designing new protocols for zero knowledge proofs, designing new consensus algorithms, designing new signature schemes, doing analysis of existing algorithms to verify whether or not whether or not they're safe, right? These are things that benefit everyone who is part of not even one single watching community, but just part of the crypto space as a whole and even going beyond the crypto space.

But these are not things that you can monetize, right? Because like once you release it, it's out there. Anyone can go and use it. If you try to keep it secret, then well, because blockchains are a space that depends so much on openness for verification, that's just going to completely ruin the trust properties of the system. Um you know, if you try to patent it, that also like you know, completely ruin people's ability to effectively build on top of your ecosystem and feel comfortable joining it.

So, it's something that is very important, but that can't be monetized using kind of you you traditional uh, like kind of capitalist you know, more market means, right? And so, the question is, well, how do we fund these things? And there's a lot of different ideas uh for how to do this. So, there's obviously centralized foundations and centralized grant-giving organizations. Um, and then there is these more interesting democratic approaches like quadratic funding is a big example, right?

Quadratic funding is basically this matching scheme where anyone can donate to projects that they find valuable, and then the system provides matching from a matching pool, right? And the way that the matching works is that it's not a fixed ratio. The amount of matching actually depends both on the amount that people contribute, and also on the number of distinct people that can contribute, right? So, if project A gets $500 from uh me and $500 from Kartik, and that's it, it might get a little bit of a match, but then if project B gets $1,000 where it's one person $1 each coming from a thousand different people, then like that good is much more public than the good that only matters to me and Kartik, and so that like that good gets a much more of a public good, right? Um, so that's this is a mechanism that's been tried out by Gitcoin Grants for a couple of years now, and it's funded a lot of interesting stuff.

Um, and another interesting one is retroactive funding, which is being experimented on by Optimism right now, where basically, yeah, instead of giving grants to things ahead of time, you or you give uh rewards to projects that have already produced some results. And there the idea is that, well, it might be easier for the mechanism to judge things that have already given good results than it is to judge uh things that are um, useful ahead of time. And And the idea is that then the retroactive grants ecosystem could be combined with a proactive investment ecosystem where people would like basically be willing to invest in projects that like they could be promising public goods in exchange for getting a share of the retroactive reward when that ends up happening, right? Um so, basically, yeah, you know, the the idea is to try to like create this pipeline for rewarding the production of public goods that it it comes as close to the quality of the pipeline that we have today for rewarding the production of private goods. Great.

And what are some of the experiments that you think So, so there's definitely some of the retroactive experiments that you described right there, but do you think that there are other interesting mechanisms that maybe could couple to whether whether it's quadratic voting or some other structures for how to let large communities orient the the allocation of funds? I mean, we've seen some good experiments in terms of definitely funding you know, things like the Gitcoin grant rounds and so on. But we haven't yet seen you know, some kind of larger scale utility or some larger scale development process being funded through public goods. What are your thoughts there? Like what are we sort of missing?

What kind of mechanisms to kind of aggregate our thinking or aggregate our preferences or kind of route the funding or collect the funding? What do you think is are some of the experiments that might kind of open up like larger scale aggregations? Mhm. I mean, I think in this case one of the big bottlenecks is of course that like unlike much of the rest of the crypto space where a lot of people easily can jump in because it's an opportunity to get out money, public goods funding is ultimately a sink for money. Like you have to the money has to come from somewhere, right?

So, in the case of Gitcoin grants, they relied on a lot of individual donations for a while, and then only more recently there have been protocols that have been contributing to the matching pools, and there have been uh like dog coins given to them obviously and things like that. But and then in the case of optimism, their retro funding is going to directly get a share of the token of the OP tokens, right? Um, but like in order to find public create these larger scale sources of funding public goods, like you have to find the larger sources of funding, right? And so, if a version of public goods funding succeeds at at a $10 million budget level, then like if you want to scale up to a $10 billion, that doesn't just happen automatically because more people get excited and want to join. Like you have to actually find the $10 billion per size matching pools, right?

And that's hard, right? And uh I think it is valuable for crypto projects to and eventually outside of crypto organizations, like corporations or governments to contribute to matching pools, but like that's an investment that's required on their part, right? And that's uh something where they would need to see the value in order to contribute to the these uh matching pools. And it's something where the value of these mechanisms and the ability of these mechanisms to scale is just something that needs to keep being proven, right? And uh I think uh the existing experiments have done a good job of proving them at small scales, but like it's uh it is something that is just inherently going to take a longer amount of time for people to feel comfortable than for, you know, some uh project where it's just like some DeFi money-making thing and uh you know, if it's attractive to people, then it can just like jump up to a $5 billion size relatively quickly, right?

It's just uh like find finding public goods funding is more meaningful than those things, but it's also just like inherently a longer slog because of the commitment that it requires. Yeah. We talk you kind of brought up a an interesting kind of subtopic here in that answer. And uh this was a tweet, I think, 2 days ago, and I quote, "The notion of governance rights as a narrative for why a token should be valuable is pathological." Uh and it followed no other second explanations or or comments or replies.

Uh, could you give us a little bit more context on well, why what this tweet actually implies and and more specifically, what is a better framework of thinking about this? Um, sure. So, I think there's this narrative that a lot of people use where they basically say that it's okay to use a valuable because it's a governance token and like the governance power is the thing that makes the token valuable, right? And like I've always thought that this narrative is extremely screwed up basically because what that narrative like in general voting is not something that's like very monetarily useful to individual personally, right? Like voting actually is an example of something that is a public good.

The reason is that every individual voter, they're making decisions that where they only have a very small probability of affecting the outcome. And the outcome that they affect matters not just for them, but it also matters for a very huge number of people, right? And if you try to translate voting into a token context, you get this effect where if you're just a small token holder, then your probability of affecting the outcome is tiny and your ability to benefit from the outcome is tiny because you only have a tiny share of the tokens and so your like actual ability to gain from voting is tiny squared, right? But then if you look at big funds that can afford to buy you know, 5 or 10% of the entire token supply of something, well, they have a big ability to influence the outcome and also their um, ability to benefit from a particular outcome is also much larger, right? Um, and so you have this kind of quadratic effect.

And so if the only value of a token is basically governance power, then like really the only people who are actually going to get enough governance power to be to be willing to pay for it are for these really large funds, right? And so you're basically saying, well, this token is valuable because of the possibility that at some at some point down the line its governance are going to get captured by by extremely rich people that will do things to kind of push the system toward their own interests, right? So, like that's not really a healthy narrative and like actually we don't really want protocols that are captured capturable by wealthy special interests, right? We want protocols that kind of serve all of their users in a in more equal way and that And so what I think we need is like we need healthier and kind of more sustainable narratives for like why protocols have value, right? And like there are I think there are definitely are narratives that actually are saying right like there's you know direct use value like especially the you know, in the case of Ethereum like you know, you have ETH and you have ETH when you send ETH to pay a transaction, the the the biggest portion of that transaction fee gets burned and so that ETH actually gets consumed to pay for the right to get it the transaction included, right?

So that's like a kind of predicting future cash flows. There's like there are better these better these better narratives that are I think like much healthier, but the problem is that's like it's harder to create a project that actually like has them, right? And like I understand are attracted to these kind of narratives for like Basically one of the unfortunately like pathological thing that happens is that these moon narratives are like more appealing for regulatory reasons because it's like the good narratives are the narratives that put you into a highly regulated category more often and the bad narratives are the narratives that like like basically if you if you just say something that something is a meme coin, you're not regulated at all, right? And so there's this kind of unfortunate effect and like I think ultimately in the long term the only solution to this is going to be to like actually talk with regulators, possibly starting in places outside of the US that are more flexible to just to kind of figure out and set examples that work. Um and then you know and you know, just smaller countries in general.

And then moving up to you know, both India and the US over time, right? But that's going to be a long and complicated discussion. It'll require buying you know, from regulators and the crypto community. But I think uh you know, in the in the short term, like it's uh it is something that I think people needs to be like more cognizant of, right? Because the this current approach of like saying things are governance tokens, like it doesn't actually uh give people the ability to distinguish to distinguish between tokens that actually have a possibility of making sense versus things that are all obfuscated meme coins, right?

And I think in general, in one of the things that the crypto space needs to get better at is the crypto space needs to get better at like distinguishing between sensible projects and scams. And that's um And so um you know, giving like being having better ways of talking about valuation narratives for tokens that like really create a big distinction between a token that's valuable because it deserves to be valuable versus a token that's valuable because it's just a meme coin bubble and it's a a meme coin bubble that's pretending to be something else to try to get more people buying it. Like that's something that we do have to get better at distinguishing between. And and speaking of governance and governance rights, um right now we've seen DAOs develop kind of direct democracies and maybe some amount of quadratic voting to start making some decisions. Uh but we haven't seen um more sophisticated or or many sophisticated decision-making structures.

And it seems right now that DAOs are you know, kind of very labor intensive for a few people that end up being, you know, very intense community managers and so on and help guide a lot of the community. And so then people end up with kind of DAO burnout and so on. And we haven't quite evolved the kind of like work structures that maybe, um, you know, governments and corporations and other kinds of structures, um, evolved. Uh, how do you think this is going to play out over the next, uh, couple of years? Like do you think we should be make doing more experiments in kind of like, um, dividing authorities for DAOs or, um, you know, how do how do we like build a a an open governance structure that doesn't kind of fully centralize into hierarchies, but at the same time it it's able to kind of coordinate better in a larger scale?

I mean it feels like, uh, yeah, from direct democracy to republic is kind of the phase we're at and we're not sure if it goes back to the models we have now versus a a new unique one that is maybe more interesting and possible. Yeah, I mean I think there's, uh, going to be the this, uh, process over the next few years where we really figure out what is the difference between a DAO that makes no sense and a DAO that does make a lot of sense, right? Um, so, like one of the posts that I wrote a a few months ago is a DAOs are not corporations and I basically argued that the class of DAO that tries to be a company is probably not something that actually makes any sense, right? And for companies, uh, it might even be better for companies to just be regular companies and, um, you know, if they want to uh, do things like give more assurances to their users, you know, they could, uh, there's other kind of non-DAO ways of doing that like committing to open source things, for uh, for example, and, uh, you know, a a lot of kind of innovation within traditional structures. Like basically the, uh, difference is just, uh, like the difference between Microsoft being a company and Microsoft being a DAO where the DAO governance structure is the same as what the company would be is not large, right?

Um, and so we have to like separate between that and the cases where DAOs are really meaningful. And I think there's two cases where generally DAOs can be meaningful. One of those cases is like basically things that need to go very fast, and the other is things that need to go very slow, right? And so things that need to go very fast are like, you know, small groups of people coming together possibly from across the world for some specialized task, and you know, having a high level of um agility is supporting like lots of different groups working in uh working in different ways. Um so that kind of uh thing is uh something that is uh like DAOs can actually just be more efficient than uh traditional corporate structures for doing this.

DAOs can provide a greater level of decentralization, a greater level of trustlessness, um and like, you know, they just work in the kind of the light overhead nature of them actually works well. And then the other extreme is DAOs that are intentionally slow to change, right? So, if you want to have a stable coin run by a DAO, for example, one of the challenges is that the market cap of the stable coin is going to be much larger than the than the uh expected, you know, discounted future profits that the stable coin system can extract, right? And so, the stable coin and the assets that are collateralized by the system are going to be much larger than the market cap of the profit token. And so, if you have token voting governance, then it's just inevitably open to this attack where somebody buys up half of the profit token and then uses that asset basically attack and take over the system and uh grab up a huge portion of the underlying assets, right?

Um and so, you need governance that is kind of intentionally slow, that intentionally adds friction, is intentionally non-financialized in a lot of ways, possibly gives people an opportunity to leave if the system starts being taken over and all of these kinds of things, right? And like that that kind of governance is a a governance that is intentionally not efficient and intentionally frustrating in a lot of ways and its only goal is to just survive, right? Like one of the things that you notice in crypto after being here for a long time is that like one of the ways to win in the space is to just survive, right? Like I remember on Twitter there was this like I think autism capital or like one of those they made this long list of you know quote main characters in crypto. Like basically people who are like you know big crypto personalities and then like basically we checked well how many of them are still around and how many of them have suffered some kind of economic downfall or just you know disappeared into a luxury Caribbean retirement and the number of X's I think was a larger than number of check marks, right?

And a lot of those X's were pretty spectacular, right? So you know you have Mount Gox, you have you know now you have Luna, you have FTX, you have a a lot of and EOS is one that kind of you know I think faded into the background quietly but at the same time you know people are like there was Block.one that just kind of made billions of dollars off of it and is you know just you know wildly taking a you know going on a vacation somewhere, right? And so just you know creating a system that survives for a long time and continues to be aligned for a long time and particularly that solves what some of your also the succession problem, right? Like the problem of a creating a system that continues to serve particular values even after the original founders that put those values in place are gone.

Like that's hard, right? And so in like DAOs that aim for stability the hardest work is not even in the code cuz you have the code the hardest work is figuring out exactly what are the rule what should the rules be that the code tries to implement. And that's like a big and a challenge and I think things like Gitcoin grants and optimism and stablecoins, they're all going to need going to need to solve that. And if they can't solve that, then they've obviously created something that's really great, right? But they have to actually get there.

Yeah, it seems like the next output is a lot more experimentation and then see kind of how do we get these communities to actually persist and grow before we kind of settle on a model that actively works over a longer cohort size instead of just uh a little while. Maybe a quick one uh that really is a a recurring theme especially at a global events. Uh there is a really big shortage of Goerli testnet ETH. How do we make sure the future testnets don't get into the same situation? Mhm.

That's a good question. I don't know. I feel like the the and if the core devs that have thought a lot about testnets probably have better ideas than I do, but like I think in general like what could you do? You could just pre-mine a huge amount of ETH and uh put that into some organization that just like gives it out on a faucet and uh like keeps enough so it can so the faucet is able to give keep giving more testnet ETH to people who need it. tim.

beiko@ethereum.org. Just just email that guy and he'll send you some ETH. Uh so there's a lot of interest in social networks and you know you and I have talked a lot about we together the three of us have talked a lot about social networks um over over the years and we are seeing a lot of activity happening around the world with you know balkanization of social networks. There's of course all the Twitter madness recently.

Um people want to move to other more decentralized structures uh and yet we haven't quite been able to crack the nut on making uh social networks work well. There's some like really good highlights like Lens and others uh who are like um sort of like getting there. What do you think is holding us back? Like what what are the primitives that are missing? Um what's going on?

Like, why haven't we been able to make a really strong social network that's Web3 Web3 first? I think the thing that's holding back social decentralized social networks is the same thing that's holding back network states, which is that a lot of the people think of these as it being a technology play or even a social technology play. But, I think ultimately the successful ones are going to be a community play, right? Like, the the problem that all of these things have is they have to fight against a really powerful network effects gradient, where at the beginning they have you know, very few people competing against the systems that have very many people. And the only way to survive in that kind of environment is to have a high-quality people, right?

And I think so, trying to figure out like how to have a actually good and a high-quality community from the start in these systems so that people would want to use them despite the lack of scale. Like, that's a really big challenge. Um, so far it feels like a Orchester has done a better job than many others from what I can tell. Like, I've hooked into it and I've used it a bit and like, you know, discussions on there seem high quality. It's not a ghost town, right?

And they've succeeded at that. Um, so yeah, I think the question is how like, if you're building one of these, like, how do you actually create a community that's people want to participate in and where that community has like a lot of interesting people that enjoy being there with each other and they can attract other interesting people over time as well, right? And uh I don't think like that's that's not the lens through which the uh failing projects are considering this through, right? And I think that is the way that successful projects are is are considering it through. Absolutely fair.

Looks like time time will tell or you got to really force everybody to sign up for that account. Um We got a couple questions on this theme. So, I'll kind of combine all of them and sort of synthesize it to a generalized question. But, kind of from your perspective, what's one of the biggest dangers that the Ethereum ecosystem is facing? What's one challenge that the Ethereum ecosystem is facing?

Yes. Mhm. Um I think there's like basically the biggest one is this kind of implicit race between scaling technology and the kind of impending next moment of large scale adoption, right? Whether like that's you know a kind of bubble in the same way that we've had previous bubbles or just you know any large increase of people wanting to use the chain in general. Like things are in a bit of a lull right now where like I was just you know bit in the mouth before that you can include the transaction before 14 cents sometimes.

But, that's that wasn't the status quo a year ago and that wasn't the status quo two years ago, right? One to two years ago we had a crazy fee market and it would take multiple dollars to send even a simple transaction. And then if you want to send the ERC-20 you would have to pay over 10. If you want to send somebody ZK-Snark privacy preserving thing you'd have to pay hundreds of dollars and it was just completely insane, right? Um so, I think right now we've been given this kind of temporary respite from that situation.

But, like that's temporary, right? Or rather if it ends up not being temporary then the then crypto has failed, right? And so, we have to scale fast enough to actually be ready for that next moment and to make sure that the transactions in the ecosystem actually continue to be affordable. And that's not something that's happened yet. And that's something where obviously roll-up teams and a lot of bunch of other people are really really working hard on, but that's something that we uh uh you know, we need to keep working hard on to make sure that we uh actually get more scaling by the time that we need it.

Um So, we're here in the India right now with uh a couple thousand people. About I think 20,000 people applied try to get into this hackathon. Um the Web3 community is growing a ton in India uh and is one of the most populous countries in the world, one of the largest economies. The most populous country in the world, one of the largest economies. Um there's a lot of builders.

Uh there's a lot of growth in the ecosystem. Uh people are are super excited to build uh really great use cases and really great things. What's some of the advice that you have for builders uh setting off in their Web3 journey, starting to build their applications, whether it's kind of like the things to try and build or kind of um how to kind of set a vision and stay the course or, you know, maybe looking back in your own um development, like what are the things that you um ended up kind of thinking were really important as a builder um to kind of like share with the community here so they, you know, you can basically influence a lot of people over the next couple of years in building pretty great things. Mhm. Yeah, I think uh the big theme this decade for crypto is just is going to have to be build things that don't that aren't just like fascinating and cool in theory, but that actually really work for people in practice, right?

Um so, understand what people's specific needs are um even if I, you know, people's needs are just fairly stuff that seems fairly simple and boring, like um you know, ways of holding crypto stable coins and uh you know, moving moving stable coins around. Um try to you know, really like I think in the last decade there was a lot of building out of theory and there was a lot of uh people thinking through, well, okay, you know, we could have prediction markets, so we could have uh you know, DAOs and uh we could have games and NFTs and we could do fancy stuff. And uh a lot of this kind of very theoretical analysis of what kinds of things we could do, which was great and I think very appropriate for the early stage of the space, but like now what we really need is like actually build things that are going to be meaningful for you know, many millions of users, right? And I think India is a great place to start working on that. I think basically anywhere other than existing rich countries is a great place to work on that because that's where the a lot of the meaningful problems are, right?

One of the things I've noticed is that the DeFi space in you know, places like New York, like you know, it does often basically end up trying to like really focusing a lot on the needs of a fairly small number of wealthy people to try to you know, gamble or like earn like 11% instead of 6% on yield farming or whatever, right? And in ways that are in the long term not super sustainable. And I think in like for for example, like in Latin America where I have spent a lot of a lot of time this year in particular and like you know, you see people that have a much more kind of real attitude and you know, and that understand that you know, hey, there's serious problems that we're working on and that that try to build things that are useful for people for people who have these really serious problems and you know, I imagine you know, India obviously has a its own you know, very complex environment and I think this this combination of a lot of challenges and a lot of very smart people that are willing to meet those challenges and that's in you know, an opportunity for the for people building in crypto to build things that you know, are actually going to really really matter for people's lives. So yeah, I guess you know, biggest advice is that it is to focus more on the users and like focus more on the users and like really think through what are the things that are going to be useful for people and not just in the next 6 months of a bull cycle, but you know, in the next 10 years. It's Alec, that is a amazing positive note to to end this chat on.

I want to thank you so much for joining us and I hope to see you in person for the next one. Thank you so much. Yeah, no, I'm definitely very much looking forward to coming in to you in person. Amazing.